Research on the Optimal Pricing Strategy of Supermarket Vegetables Based on Elastic Network Regression Model

Abstract

This study develops optimal restocking and pricing strategies for vegetables in supermarkets to reduce significant post-harvest losses in China, which currently result in losses exceeding 100 billion yuan annually. By employing cost-plus pricing, ridge regression, and support vector regression (SVR), this research analyzed the relationship between pricing and sales volumes for various vegetable categories. Predictive models for the week of July 1-7, 2023, indicated a consistent markup rate of about 58%, confirming the reliability of the models. For single-item analysis, both a greedy algorithm and elastic net regression were used. The greedy algorithm corrected loss rates and sales quantities to maximize profit, while the elastic net regression addressed multicollinearity and overfitting issues, leading to more accurate sales predictions. Integrating these approaches effectively predicted sales and optimized supermarket restocking and pricing strategies. The findings showed significant improvements in expected profits, validating the effectiveness of these combined models. This research demonstrates the potential for advanced analytical techniques to enhance supermarket operations, reduce waste, and better meet consumer demand.